Name
Abstract | ||
|---|---|---|
Denoising is a critical preconditioning step for quantitative analysis of medical images. Despite promises for more consistent diagnosis, denoising techniques are seldom explored in clinical settings. While this may be attributed to the esoteric nature of the parameter sensitve algorithms, lack of quantitative measures on their efficacy to enhance the clinical decision making is a primary cause of physician apathy. This paper addresses this issue by exploring the effect of denoising on the integrity of supervised lung parenchymal clusters. Multiple Volumes of Interests (VOIs) were selected across multiple high resolution CT scans to represent samples of different patterns (normal, emphysema, ground glass, honey combing and reticular). The VOIs were labeled through consensus of four radiologists. The original datasets were filtered by multiple denoising techniques (median filtering, anisotropic diffusion, bilateral filtering and non-local means) and the corresponding filtered VOIs were extracted. Plurality of cluster indices based on multiple histogram-based pair-wise similarity measures were used to assess the quality of supervised clusters in the original and filtered space. The resultant rank orders were analyzed using the Borda criteria to find the denoising-similarity measure combination that has the best cluster quality. Our exhaustive analyis reveals (a) for a number of similarity measures, the cluster quality is inferior in the filtered space; and (b) for measures that benefit from denoising, a simple median filtering outperforms non-local means and bilateral filtering. Our study suggests the need to judiciously choose, if required, a denoising technique that does not deteriorate the integrity of supervised clusters. |
Year | DOI | Venue |
|---|---|---|
2012 | 10.1117/12.911650 | Proceedings of SPIE |
Keywords | Field | DocType |
Denoising,supervised classification,cluster validity,Borda count,non-local means | Noise reduction,Cluster (physics),Data mining,Histogram,Borda count,Clinical decision making,Pattern recognition,Non-local means,Artificial intelligence,Medical diagnostics,Physics | Conference |
Volume | ISSN | Citations |
8315 | 0277-786X | 0 |
PageRank | References | Authors |
0.34 | 2 | 6 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| padmapriya jayamani | 1 | 0 | 0.34 |
| Sushravya Raghunath | 2 | 1 | 2.85 |
| Rajagopalan Srinivasan | 3 | 683 | 79.21 |
| Ronald A Karwoski | 4 | 24 | 7.43 |
| Brian J Bartholmai | 5 | 33 | 11.91 |
| Richard A. Robb | 6 | 645 | 238.12 |